Pattern Grouping Method Based on Machine Learning

4. The method of claim 2, wherein the defect patterns comprise GDS information associated with defects.

5. The method of claim 4, wherein the defect patterns comprise information derived from the GDS information that includes number of sides, number of angles, dimension, shape, or a combination thereof.

6. The method of claim 1, wherein the fixed-dimensional feature vector is a one-dimensional feature vector.

12. The system of claim 11, wherein the pattern data comprises a sample design plan.

14. The system of claim 13, wherein the plurality of defect patterns is retrieved from a storage.